Skylight integrated smoke intelligent sensing control system

Abstract

The invention discloses a skylight integrated flue gas intelligent sensing control system, and relates to the technical field of automobile safety, and the system comprises a flue gas intelligent sensing module which carries out the judgment of the flue gas state of parameters based on a trained flue gas state judgment model; the flue gas state analysis module is used for further analyzing the flue gas state based on the trained flue gas state analysis model; the skylight driving control module is used for controlling the running state of the skylight; the skylight operation state monitoring module is used for analyzing the operation state data of the skylight; and the early warning module is used for triggering an early warning signal of a corresponding level based on a preset multi-level early warning mechanism. According to the invention, an intelligent system which integrates active smoke exhaust control, mechanical state closed-loop management, multi-dimensional early warning and vehicle-specification-level environment adaptation is constructed by fusing a plurality of algorithms according to the characteristics of the closed space of the automobile, so that the whole-process upgrading of the automobile sunroof from passive response to active prevention and control is realized; and the accuracy and safety of automobile smoke prevention and control are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of automotive safety technology, specifically to a sunroof integrated intelligent smoke sensing and control system. Background Technology

[0002] As one of the comfort features of modern cars, the sunroof is widely used in mid-to-high-end models. Its main function is to improve ventilation and lighting inside the car and enhance the comfort experience of drivers and passengers. With the development of technology, the traditional manual sunroof has been gradually replaced by electric sunroof, and the control method of electric sunroof has also developed from simple on / off control to intelligent control. Currently, most car sunroof control systems rely on simple on / off operations, with users controlling the sunroof's opening, closing, or angle adjustment via buttons. However, this traditional control method fails to consider changes in the external environment, especially in extreme conditions such as smoke and weather. In these situations, the sunroof's automatic control and safety are not adequately guaranteed. In recent years, with the development of intelligent sensing technology, artificial intelligence, and big data technology, intelligent perception control systems have been increasingly widely used in the automotive field. Particularly in scenarios such as smoke leaks and fire warnings, intelligent perception systems can monitor changes in the environment inside and outside the vehicle in real time and automatically determine whether to open the sunroof to ensure the safety of the occupant. However, the market still lacks systems that integrate intelligent smoke sensing and sunroof control functions, which limits the safety and intelligence level of sunroofs under extreme conditions.

[0003] Therefore, it is necessary to develop an integrated system based on intelligent smoke sensing and automatic sunroof control, which can automatically adjust the sunroof status according to changes in the environment inside and outside the vehicle. This can improve both the driver's comfort and safety, especially in the event of dangerous situations such as smoke or fire, and can respond in a timely manner and take automatic measures to protect the driver's safety. Summary of the Invention

[0004] The purpose of this invention is to solve the problems mentioned above by proposing an integrated intelligent smoke sensing and control system for sunroofs.

[0005] The objective of this invention can be achieved through the following technical solutions: The sunroof integrated smoke detection and control system includes: Intelligent flue gas sensing module: It is used to collect relevant parameters of flue gas in the monitoring area in real time. Based on the trained flue gas state judgment model, it judges the flue gas state of the parameters. If the flue gas state is safe, it is transmitted to the flue gas state analysis module for analysis and prediction. If the flue gas state is dangerous, it is transmitted to the early warning module for flue gas early warning. Flue gas state analysis module: Receives flue gas state judgment information transmitted by the flue gas intelligent sensing module, and further analyzes the flue gas state based on the trained flue gas state analysis model to obtain flue gas state data; Sunroof drive control module: Based on the current flue gas status data and the sunroof operating status, it generates corresponding control commands based on preset control logic to drive the sunroof to perform opening, closing and angle adjustment actions, and control the sunroof operating status; Sunroof Operation Status Monitoring Module: Collects real-time sunroof operation status data, analyzes the sunroof operation status data based on a trained sunroof anomaly analysis model, and outputs the sunroof operation status. Early warning module: Receives hazardous flue gas status information and skylight operation status information transmitted by the intelligent flue gas sensing module, and triggers corresponding level early warning signals based on a preset multi-level early warning mechanism.

[0006] Preferably, the intelligent flue gas sensing module specifically includes: Sensing parameter acquisition unit: Based on sensors, it collects real-time data on flue gas concentration, ambient temperature, CO concentration and relative humidity within the monitoring area at a frequency of once per second, and adds timestamp and spatial coordinate information to each parameter; Data preprocessing unit: Employs a moving average filtering algorithm to filter and denoise the acquired raw parameters, based on 3D... Criteria for removing outliers; Flue gas state judgment model unit: It adopts a trained random forest classification model, takes the pre-processed flue gas concentration, temperature, CO concentration and humidity as input features, and outputs the flue gas state judgment results of safety and danger; The model training samples include 10,000+ sets of historical flue gas data. The safety status judgment thresholds are: flue gas concentration < 500 ppm, temperature < 50℃, and CO concentration < 200 ppm.

[0007] Preferably, the flue gas state judgment model unit specifically includes: The random forest classification model of the flue gas state judgment model unit contains 50 decision trees, each with a maximum depth of 8 layers, and uses the Gini coefficient as the feature importance evaluation index.

[0008] Preferably, the flue gas state analysis module specifically includes: Temporal feature extraction unit: performs temporal analysis on the safety status parameters transmitted by the flue gas intelligent sensing module and extracts the average value of the parameters within 10 minutes; Flue gas state analysis model unit: Using a trained LSTM neural network model, with time-series features as input, it outputs the predicted values ​​of flue gas concentration, temperature, and diffusion trend vector for the next 5 minutes, forming flue gas state data; Trend warning subunit: If the predicted flue gas concentration is ≥400ppm, a potential risk warning is generated and transmitted to the sunroof drive control module to trigger the sunroof pre-opening preparation in advance.

[0009] Preferably, the sunroof drive control module specifically includes: Control logic storage unit: Presets multiple sets of mapping rules between flue gas status and sunroof operation, including: When the flue gas concentration is 500-1000ppm and the temperature is 50-60℃, control the skylight to open to 30°. When the flue gas concentration is 1001-2000ppm and the temperature is 61-70℃, control the skylight to open to 60°. When the flue gas concentration is greater than 2000 ppm and the temperature is greater than 70°, the skylight should be opened to 90°. Once the flue gas condition returns to a safe level and remains so for 3 minutes, control the skylight to close to 0°. Drive execution unit: It consists of a servo motor, a reduction gear set and a limit switch. After receiving the control command, it drives the sunroof to adjust the angle at a speed of 5° per second. The angle control accuracy is ±1°. It also has an overload protection function. When the motor current is >5A, it will automatically stop. Status feedback unit: Real-time acquisition of sunroof current angle, motor operating current and action completion status, and feedback to sunroof operation status monitoring module.

[0010] Preferably, the control logic of the sunroof drive control module further includes: Emergency Priority Subunit: When both a smoke hazard signal and a sunroof malfunction signal are received simultaneously, the forced opening command is executed first. If the target angle is not reached within 30 seconds, a Level 3 warning is triggered. Linkage control subunit: Supports linkage with the building fire protection system. When receiving a fire linkage signal, it forces the skylight to open to 90° and locks it in manual control mode.

[0011] Preferably, the sunroof operation status monitoring module specifically includes: Operational data acquisition unit: Collects sunroof motor current, actual opening angle, vibration frequency, and action response time through current sensor, angle sensor, and vibration sensor; The sunroof anomaly analysis model unit uses a pre-trained SVM anomaly detection model, takes the running data as input, and outputs the running status of normal, slightly abnormal, and severely abnormal. Based on the abnormal operation status of the skylight, a severe anomaly judgment condition is used to determine the severe anomaly; Abnormal log recording unit: Automatically records the time of occurrence, duration and associated parameters of abnormal states, forming a device health record, and triggers a device maintenance reminder when three serious abnormalities are accumulated.

[0012] Preferably, the criteria for determining severe anomalies specifically include: Abnormal motor current: The sunroof drive motor has an operating current of >8A and lasts for ≥2 seconds, and the current fluctuation exceeds ±50% of the rated current of 5A for 3 consecutive sampling cycles. Abnormal mechanical action: After receiving the drive control command, the actual opening angle of the sunroof deviates from the target angle by more than 5° for 5 seconds, and the response time for 3 consecutive actions is more than 3 seconds. Abnormal structural condition: The vibration frequency detected by the vibration sensor is >20Hz and the amplitude is >0.5mm, and the angle sensor feedback value does not change within 10 seconds; Communication link abnormality: Communication interruption time with the sunroof drive control module > 10 seconds, 5 consecutive data transmission failures, and data packet loss rate > 30%.

[0013] Preferably, the skylight anomaly analysis model unit specifically includes: The SVM anomaly detection model of the sunroof anomaly analysis model unit uses normal operation data as training samples, including 8000+ sets of sunroof fault-free operation records, and maps the data to a high-dimensional space through kernel functions; An anomaly detection hyperplane is constructed, with the anomaly determination threshold set to an average distance from the normal sample set greater than 0.8. The formula for calculating the average distance is as follows: ; In the formula, This is the normal sample size. The sample to be tested. For the first A normal sample, This represents the average distance.

[0014] Preferably, the early warning module specifically includes: Based on the combination of the degree of smoke hazard and the operation status of the skylight, the warning levels are divided into: Level 1 Warning: The flue gas condition is dangerous, with a concentration of 500-1000 ppm, and the skylight is operating normally; Level 2 warning: The flue gas condition is dangerous, with a concentration of 1001-2000 ppm, and the skylight shows slight abnormalities; Level 3 warning: The flue gas condition is dangerous, the concentration is >2000ppm, and the skylight is severely abnormal; Early warning execution unit: Triggers corresponding measures for different levels, including: Level 1 Warning: Activate local audible and visual alarms and push information to the on-site monitoring screen; Level 2 warning: In addition to Level 1 warning, an alarm signal is sent to the fire control room, and the make-up air device in the adjacent area is activated in conjunction with it; Level 3 warning: In addition to Level 2 warning, it triggers the fire emergency broadcast, automatically dials the preset fire alarm number and the mobile phone of the management personnel, and simultaneously uploads on-site video footage.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. In response to the characteristics of rapid smoke diffusion and complex environment in the enclosed space of a car, a multi-dimensional sensor integrating smoke detection, temperature, CO concentration and humidity is integrated. High-frequency acquisition once per second and moving average filtering algorithm are used to reduce parameter noise. Combined with random forest model to classify preprocessed data, the accuracy of smoke state judgment is improved to more than 95%, solving the problem of high false alarm rate of traditional car smoke detectors with single parameters.

[0016] 2. Innovation in Smoke Trend Prediction and Sunroof Linkage Control Based on LSTM: Introducing time-series feature analysis into the control of automotive sunroofs, using LSTM neural networks to model smoke parameters under safe conditions for 10 minutes, predicting concentration change trends 5 minutes in advance, and combining with hierarchical control logic to achieve an intelligent upgrade from passive response to active smoke exhaust, with a response speed 2 to 3 orders of magnitude faster than traditional manual opening.

[0017] 3. For special operating conditions such as vehicle bumps and vibrations, an SVM anomaly detection model is designed. By mapping a high-dimensional feature space through a kernel function, it accurately identifies serious anomalies such as motor current >8A and angle deviation >5° with an average distance >0.8 as the threshold. Combined with an anomaly log and maintenance reminder mechanism, a detection-early warning-maintenance closed loop is constructed, reducing the occurrence rate of sunroof mechanical failure by 40%. A three-level early warning system is constructed by integrating smoke concentration and sunroof status to solve the pain point of the disconnect between traditional vehicle fire protection and sunroof control, and shorten the rescue response time. Attached Figure Description

[0018] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0019] Figure 1 This is a system framework diagram of the present invention; Figure 2 This is a diagram showing the internal system framework of the intelligent flue gas sensing module in this invention. Figure 3 This is a diagram showing the internal system framework of the flue gas state analysis module in this invention. Figure 4 This is an internal system framework diagram of the sunroof drive control module in this invention; Figure 5 This is a diagram of the internal system framework of the sunroof operation status monitoring module. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0022] Please see Figure 1 As shown, the sunroof integrated smoke detection and control system includes: Intelligent flue gas sensing module: It is used to collect relevant parameters of flue gas in the monitoring area in real time. Based on the trained flue gas state judgment model, it judges the flue gas state of the parameters. If the flue gas state is safe, it is transmitted to the flue gas state analysis module for analysis and prediction. If the flue gas state is dangerous, it is transmitted to the early warning module for flue gas early warning. Flue gas state analysis module: Receives flue gas state judgment information transmitted by the flue gas intelligent sensing module, and further analyzes the flue gas state based on the trained flue gas state analysis model to obtain flue gas state data; Sunroof drive control module: Based on the current flue gas status data and the sunroof operating status, it generates corresponding control commands based on preset control logic to drive the sunroof to perform opening, closing and angle adjustment actions, and control the sunroof operating status; Sunroof Operation Status Monitoring Module: Collects real-time sunroof operation status data, analyzes the sunroof operation status data based on a trained sunroof anomaly analysis model, and outputs the sunroof operation status. Early warning module: Receives hazardous flue gas status information and skylight operation status information transmitted by the intelligent flue gas sensing module, and triggers corresponding level early warning signals based on a preset multi-level early warning mechanism.

[0023] Please see Figure 2 As shown, the intelligent flue gas sensing module specifically includes: Sensing parameter acquisition unit: Based on sensors, it collects real-time data on flue gas concentration, ambient temperature, CO concentration and relative humidity within the monitoring area at a frequency of once per second, and adds timestamp and spatial coordinate information to each parameter; Data preprocessing unit: Employs a moving average filtering algorithm to filter and denoise the acquired raw parameters, based on 3D... The criteria for removing outliers include the moving average filtering algorithm, which is formulated as follows: ; In the formula, After moving average filtering, the first The parameter values ​​of each sampling point The size of the sliding window in this system =5, meaning that each filtering calculation includes 5 data points: the current sampling point and the previous 4 sampling points. The summation index takes values ​​from 0 to n−1. For the first The original parameter values ​​of each sampling point; Flue gas state assessment model unit: This unit employs a pre-trained random forest classification model, using pre-processed flue gas concentration, temperature, CO concentration, and humidity as input features. It outputs a judgment result indicating whether the flue gas state is safe or hazardous. The random forest model output formula is as follows: ; In the formula, For the number of decision trees, This is the output of the first decision tree. This is the output of the second decision tree. The mode function is used to select the category that appears most frequently from the outputs of multiple decision trees as the final output of the model. The final flue gas state judgment result output by the random forest classification model is either safe or hazardous. The model training samples include 10,000+ sets of historical flue gas data. The safety status judgment thresholds are: flue gas concentration < 500 ppm, temperature < 50℃, and CO concentration < 200 ppm. This intelligent flue gas sensing module, by introducing a moving average filtering algorithm and a random forest classification model, provides a comprehensive judgment on multiple parameters such as flue gas concentration, temperature, and CO concentration. It can accurately distinguish between safe and dangerous flue gas states, while ensuring high accuracy and efficiency in data acquisition.

[0024] The flue gas state judgment model unit specifically includes: The random forest classification model of the flue gas state judgment model unit contains 50 decision trees, each with a maximum depth of 8 layers. The Gini coefficient is used as the feature importance evaluation index. The formula for calculating the Gini coefficient is as follows: ; In the formula, For the first The probability of a class A smaller value indicates higher dataset purity and a more concentrated class distribution. The total number of categories, Of these, flue gas concentration accounted for 40% of the weighting, CO concentration for 30%, temperature for 20%, and humidity for 10%. Gini coefficient; The optimized configuration of the random forest model, with each decision tree having a depth of 8 layers and the Gini coefficient used to measure feature importance, improves the model's classification performance and accuracy. In addition, the weighting of flue gas concentration, CO concentration, temperature, and humidity enables the model to more accurately predict flue gas conditions.

[0025] Please see Figure 3 As shown, the flue gas state analysis module specifically includes: Temporal feature extraction unit: performs temporal analysis on the safety status parameters transmitted by the flue gas intelligent sensing module and extracts the average value of the parameters within 10 minutes; Flue gas state analysis model unit: Using a trained LSTM neural network model, with time-series features as input, it outputs the predicted values ​​of flue gas concentration, temperature, and diffusion trend vector for the next 5 minutes, forming flue gas state data; Trend warning subunit: If the predicted flue gas concentration is ≥400ppm, a potential risk warning is generated and transmitted to the sunroof drive control module to trigger the sunroof pre-opening preparation in advance; The introduction of LSTM neural networks enables time-series analysis of flue gas data and prediction of future flue gas conditions, providing predictive decision support for sunroof drive control, thereby allowing for proactive measures to avoid safety hazards.

[0026] Please see Figure 4 As shown, the sunroof drive control module specifically includes: Control logic storage unit: Presets multiple sets of mapping rules between flue gas status and sunroof operation, including: When the flue gas concentration is 500-1000ppm and the temperature is 50-60℃, control the skylight to open to 30°. When the flue gas concentration is 1001-2000ppm and the temperature is 61-70℃, control the skylight to open to 60°. When the flue gas concentration is greater than 2000 ppm and the temperature is greater than 70°, the skylight should be opened to 90°. Once the flue gas condition returns to a safe level and remains so for 3 minutes, control the skylight to close to 0°. Drive execution unit: It consists of a servo motor, a reduction gear set and a limit switch. After receiving the control command, it drives the sunroof to adjust the angle at a speed of 5° per second. The angle control accuracy is ±1°. It also has an overload protection function. When the motor current is >5A, it will automatically stop. Status feedback unit: Real-time acquisition of sunroof current angle, motor operating current and action completion status, and feedback to sunroof operation status monitoring module.

[0027] The control logic of the sunroof drive control module also includes: Emergency Priority Subunit: When both a smoke hazard signal and a sunroof malfunction signal are received simultaneously, the forced opening command is executed first. If the target angle is not reached within 30 seconds, a Level 3 warning is triggered. Linkage control subunit: Supports linkage with the building fire protection system. When receiving a fire linkage signal, it forces the skylight to open to 90° and locks it in manual control mode. The sunroof drive control module integrates the mapping rules between smoke state and sunroof movement into its control logic, realizing multi-level adjustment and having the ability to automatically adjust the sunroof angle. Its precise drive execution and overload protection functions greatly enhance the intelligence and safety of the system.

[0028] Please see Figure 5 As shown, the sunroof operation status monitoring module specifically includes: Operational data acquisition unit: Collects sunroof motor current, actual opening angle, vibration frequency, and action response time through current sensor, angle sensor, and vibration sensor; The skylight anomaly analysis model unit uses a pre-trained SVM anomaly detection model. Taking the running data as input, it outputs the running status as normal, slightly abnormal, and severely abnormal. The decision function of the SVM model is: ; In the formula, The output of the SVM model is... This is a sign function that outputs 1 when the value inside the parentheses is greater than 0, indicating normal operation, and -1 when the value inside the parentheses is less than 0, indicating an error. It is used to determine whether the sunroof's operating status is abnormal. This is the data vector of the sunroof's operating status to be detected. For the first The data vector of each training sample For Lagrange multipliers, For the first Labels of each training sample For kernel function, For bias, The number of training samples. The sample to be tested With training samples Euclidean distance, The kernel function parameter is set to 0.1. It is a natural exponential function; Based on the abnormal operation status of the skylight, a severe anomaly judgment condition is used to determine the severe anomaly; Anomaly Log Recording Unit: Automatically records the occurrence time, duration, and associated parameters of anomalies, forming a device health record. When three serious anomalies are recorded, a device maintenance reminder is triggered. The sunroof operation status monitoring module collects operation data through multiple sensors and uses an SVM anomaly detection model for real-time anomaly analysis. This method improves the detection accuracy of sunroof operation status and provides precise monitoring of equipment health.

[0029] The specific criteria for determining severe anomalies include: Abnormal motor current: The sunroof drive motor has an operating current of >8A and lasts for ≥2 seconds, and the current fluctuation exceeds ±50% of the rated current of 5A for 3 consecutive sampling cycles. Abnormal mechanical action: After receiving the drive control command, the actual opening angle of the sunroof deviates from the target angle by more than 5° for 5 seconds, and the response time for 3 consecutive actions is more than 3 seconds. Abnormal structural condition: The vibration frequency detected by the vibration sensor is >20Hz and the amplitude is >0.5mm, and the angle sensor feedback value does not change within 10 seconds; Communication link failure: Communication interruption time with the sunroof drive control module > 10 seconds, 5 consecutive data transmission failures, data packet loss rate > 30%; Multiple anomaly detection conditions have been introduced, which can accurately identify various fault modes of the sunroof, such as abnormal motor current, abnormal mechanical action, abnormal structural status, and abnormal communication link, thus ensuring the efficient operation of the equipment.

[0030] The sunroof anomaly analysis model unit specifically includes: The SVM anomaly detection model of the sunroof anomaly analysis model unit uses normal operation data as training samples, including 8000+ sets of sunroof fault-free operation records, and maps the data to a high-dimensional space through kernel functions; An anomaly detection hyperplane is constructed, with the anomaly determination threshold set to an average distance from the normal sample set greater than 0.8. The formula for calculating the average distance is as follows: ; In the formula, This is the normal sample size. The sample to be tested. For the first A normal sample, Average distance; The SVM anomaly detection model is trained using a large amount of normal operation data to construct an anomaly detection hyperplane, which greatly improves the sensitivity and accuracy of anomaly detection. By setting a threshold, it is possible to determine in real time whether the operation status of the skylight is normal.

[0031] The early warning module specifically includes: Warning Level Classification Unit: Based on the combination of flue gas hazard level and skylight operation status, the warning levels are classified as follows: Level 1 Warning: The flue gas condition is dangerous, with a concentration of 500-1000 ppm, and the skylight is operating normally; Level 2 warning: The flue gas condition is dangerous, with a concentration of 1001-2000 ppm, and the skylight shows slight abnormalities; Level 3 warning: The flue gas condition is dangerous, the concentration is >2000ppm, and the skylight is severely abnormal; Early warning execution unit: Triggers corresponding measures for different levels, including: Level 1 Warning: Activate local audible and visual alarms and push information to the on-site monitoring screen; Level 2 warning: In addition to Level 1 warning, an alarm signal is sent to the fire control room, and the make-up air device in the adjacent area is activated in conjunction with it; Level 3 warning: In addition to Level 2 warning, it triggers the fire emergency broadcast, automatically dials the preset fire alarm number and the mobile phone of the management personnel, and simultaneously uploads on-site video footage; The early warning module introduces a multi-level early warning mechanism, which classifies early warning levels in detail based on the combination of flue gas status and skylight operation status, and provides corresponding response measures for different levels. This mechanism enhances the system's response speed and emergency handling capabilities, ensuring that the most appropriate measures are taken in different levels of dangerous situations.

[0032] In summary, the advantages of this invention are as follows: By intelligently sensing parameters such as smoke concentration, temperature, and humidity, the system can monitor the in-vehicle environment in real time. Once a dangerous smoke concentration is detected, the system will automatically open the sunroof and adjust its angle to help the driver ventilate and remove smoke in time, thus avoiding danger. The system can automatically adjust the sunroof opening status based on real-time data and preset control logic without manual intervention, greatly improving driving convenience and safety. Traditional sunroof control systems rely on manual judgment and operation, while this system, through automated control, effectively avoids safety hazards caused by improper operation by the driver in emergency situations. The system has a multi-level smoke warning mechanism. When the smoke concentration reaches different warning levels, it will remind the driver to take timely action through sound and light alarms, thereby further improving the driver's ability to react to dangerous situations. By intelligently adjusting the sunroof's opening angle, this system can not only more effectively exhaust smoke and ventilate, but also adjust the sunroof's opening degree according to changes in the vehicle's interior temperature, thereby optimizing the air quality and temperature inside the vehicle, improving the comfort of the driver and energy efficiency.

[0033] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A sunroof integrated flue gas intelligent sensing control system, characterized in that, Comprise: Flue gas intelligent perception module: for real-time acquisition of flue gas related parameters in the monitoring area, based on the trained flue gas state judgment model, the parameters are judged by the flue gas state, if the flue gas state is safe, it is transmitted to the flue gas state analysis module for analysis and prediction, if the flue gas state is dangerous, it is transmitted to the early warning module for flue gas early warning module; Flue gas state analysis module: receive the flue gas state judgment information transmitted by the flue gas intelligent perception module, based on the trained flue gas state analysis model, further analyze the flue gas state to obtain the flue gas state data; Sunroof driving control module: based on the current flue gas state data and the sunroof running state, based on the preset control logic to generate corresponding control instructions, drive the sunroof to execute open, close and angle adjustment action, control the running state of the sunroof; Sunroof running state monitoring module: real-time acquisition of sunroof running state data, based on the trained sunroof abnormal analysis model, analyze the sunroof running state data, output the sunroof running state; Early warning module: receive the dangerous flue gas state information and sunroof running state information transmitted by the flue gas intelligent perception module, based on the preset multi-level early warning mechanism, trigger the corresponding level of early warning signal.

2. The sunroof integrated flue gas intelligent sensing control system of claim 1, wherein, The flue gas intelligent perception module specifically comprises: Perception parameter acquisition unit: based on sensor, real-time acquisition of flue gas concentration, environmental temperature, CO concentration and relative humidity in the monitoring area, the acquisition frequency is 1 time per second, and time stamp and space coordinate information are added for each parameter; Data preprocessing unit: the collected original parameters are filtered and denoised by using a sliding average filtering algorithm, and abnormal values are removed based on the 3 criteria Flue gas state judgment model unit: adopt trained random forest classification model, take preprocessed flue gas concentration, temperature, CO concentration and humidity as input features, output safe and dangerous flue gas state judgment results.

3. The sunroof integrated flue gas intelligent sensing control system of claim 2, wherein, The flue gas state judgment model unit specifically comprises: the random forest classification model of the flue gas state judgment model unit contains 50 decision trees, the maximum depth of each decision tree is 8 layers, and Gini coefficient is used as the feature importance evaluation index.

4. The sunroof integrated flue gas intelligent sensing control system of claim 1, wherein, The flue gas state analysis module specifically comprises: Time series feature extraction unit: time series analysis of safe state parameters transmitted by the flue gas intelligent perception module, extract the mean value of parameters within 10 minutes; Flue gas state analysis model unit: adopt trained LSTM neural network model, take time series features as input, output flue gas concentration prediction value, temperature prediction value and diffusion trend vector in future 5 minutes, form flue gas state data; Trend early warning subunit: if the flue gas concentration prediction value is greater than or equal to 400ppm, generate potential risk prompt and transmit to the sunroof driving control module, trigger the sunroof to open in advance.

5. The sunroof integrated flue gas intelligent sensing control system of claim 1, wherein, The sunroof driving control module specifically comprises: Control logic storage unit: preset multiple sets of flue gas state and sunroof action mapping rules, including: When the flue gas concentration is 500-1000ppm and the temperature is 50-60℃, control the sunroof to open to 30°; When the flue gas concentration is 1001-2000ppm and the temperature is 61-70℃, control the sunroof to open to 60°; When the flue gas concentration is greater than 2000ppm and the temperature is greater than 70℃, control the sunroof to open to 90°; When the flue gas state returns to safety and lasts for 3 minutes, control the sunroof to close to 0°; The driving execution unit is composed of a servo motor, a speed reduction gear set and a limit switch, receives a control instruction, drives the sunroof to adjust the angle at a speed of 5° per second, the angle control precision is ±1°, and has an overload protection function, when the motor current is greater than 5A, the sunroof is automatically stopped; The state feedback unit: real-time acquisition of the current angle of the sunroof, the motor operating current and the action completion state, feedback to the sunroof operation state monitoring module.

6. The sunroof integrated flue gas intelligent sensing control system of claim 5, wherein, The sunroof driving control module further comprises: Emergency priority sub-unit: when receiving smoke hazard signals and sunroof abnormal signals at the same time, the forced opening instruction is executed preferentially, if the target angle is not reached for 30 seconds, the third level warning is triggered; Linkage control sub-unit: supports linkage with the building fire fighting system, when receiving the fire fighting linkage signal, the sunroof is forced to open to 90°, and is locked in manual control mode.

7. The sunroof integrated flue gas smart sensing control system of claim 1, wherein, The sunroof operation state monitoring module specifically comprises: Operation data acquisition unit: the sunroof motor current, actual opening angle, vibration frequency and action response time are collected through current sensor, angle sensor and vibration sensor; Sunroof abnormality analysis model unit: using the trained SVM abnormality detection model, taking the operation data as input, outputting normal, slight abnormality and serious abnormality operation state; Based on the abnormal sunroof operation state, the serious abnormality judgment condition is used to judge the serious abnormality; Abnormal log recording unit: automatically records the abnormal state occurrence time, duration and associated parameters, forms the equipment health file, and triggers the equipment maintenance reminder when the serious abnormality accumulates for 3 times.

8. The sunroof integrated flue gas smart sensing control system of claim 7, wherein, The serious abnormality judgment condition specifically comprises: Motor current abnormality: the sunroof driving motor operating current is greater than 8A and the duration is greater than or equal to 2 seconds, the current fluctuation amplitude exceeds the rated current 5A ± 50% and lasts for 3 sampling periods; Mechanical action abnormality: after receiving the driving control instruction, the actual opening angle of the sunroof deviates from the target angle by more than 5° and lasts for 5 seconds, and the continuous action response time is more than 3 seconds for 3 times; Structural state abnormality: the vibration frequency detected by the vibration sensor is greater than 20Hz and the amplitude is greater than 0.5mm, and the angle sensor feedback value has no change within 10 seconds; Communication link abnormality: the communication interruption time with the sunroof driving control module is greater than 10 seconds, the continuous 5 times data transmission fails, and the data packet loss rate is greater than 30%.

9. The sunroof integrated flue gas smart sensing control system of claim 7, wherein, The sunroof abnormality analysis model unit specifically comprises: The SVM abnormality detection model of the sunroof abnormality analysis model unit takes normal operation data as training samples, contains 8000+ groups of sunroof fault-free operation records, and maps the data to a high-dimensional space through a kernel function; An abnormality detection hyperplane is constructed, and the abnormality judgment threshold is set to be greater than 0.8 from the average distance of normal sample set.

10. The sunroof integrated flue gas smart sensing control system of claim 1, wherein, The warning module specifically comprises: Warning level division unit: based on the combination of smoke hazard degree and sunroof operation state, the warning level is divided into: First level warning: smoke state is dangerous, concentration is 500-1000ppm, and sunroof operation is normal; Second level warning: smoke state is dangerous, concentration is 1001-2000ppm, sunroof is slightly abnormal; Third level warning: smoke state is dangerous, concentration is greater than 2000ppm, sunroof is seriously abnormal; Warning execution unit: corresponding measures are triggered for different levels, including: Primary warning: start local sound and light alarm, push information to the on-site monitoring screen; Secondary warning: in addition to the primary warning, send an alarm signal to the fire control room, and start the adjacent area air supply device; Tertiary warning: in addition to the secondary warning, trigger the fire emergency broadcast, automatically dial the preset fire alarm phone and the manager's mobile phone, and upload the on-site video screen simultaneously.

Images